Geant4 9.5 Release Notes

December 2nd, 2011

The code and binary libraries for the supported systems are available through our Source Code Web page.

We are grateful for the efforts of Geant4 users who have provided detailed feedback or comprehensive reports of issues. We thank in particular those who have contributed corrections, improvements or developments included in this release.

Please refer to the Geant4 User Documentation for further information about using Geant4.

Contents

1. Supported and Tested Platforms
2. Supported CLHEP version
3. Items for migration of the user code
4. New Developments and Capabilities
5. Expected effects on physics and performance
6. Known Run-Time Problems and Limitations
7. Compilation Warnings
8. Known Run-Time Warnings
9. Geant4 Software License
10. Detailed list of changes and fixes

1. Supported and Tested Platforms

• Linux, gcc-4.1.2 and gcc-4.3.2 (SLC5).
  Tested on 32 and 64 bit architectures (Intel or AMD) with Scientific Linux CERN 5 (SLC5) (based on RedHat Linux Enterprise 5).
• MacOSX 10.7 with gcc-4.2.1
• Windows7 with Visual C++ 10.0 (Visual Studio 2010)

• Linux, gcc-4.6.2
• Linux, Intel-icc 11.1, 12.0
• MacOSX 10.6 with gcc-4.2.1
• Windows/XP with Visual C++ 9.0 (Visual Studio 2008)

• AIX 4.3.2, xlC 6.0
• DEC V4.0, cxx C++ V6.1-027
• HP 10.20, aCC C++ B3910B A.01.23
• MacOSX 10.4, gcc-3.3
• MacOSX 10.5, gcc-4.0.1
• SGI V6.5.5, CC 7.2.1
• SUN Solaris 5.8, C++ CC-5.5.

2. Supported CLHEP version

3. Items for migration of the user code

• The CMake system for installation replaces the old Configure script. CMAKE has been extended to cover all features. Complete information on how to use the new CMake system can be found in the User Installation Guide (Note: the old Configure script has been removed, as planned).
• Geant4 now comes with an embedded CLHEP module (based for this release on version 2.1.1.0 of CLHEP), which includes the subset of CLHEP library classes required by Geant4. On installation there is a choice between using this embedded module or installing CLHEP as an external library. Through the GNUMake system, the use/installation of the external CLHEP is triggered if the environment variable CLHEP_BASE_DIR is set; otherwise the embedded CLHEP module is used. With the CMake system, the embedded CLHEP is selected by default.
• An additional package, expat, required by the the new hadronic lend module is also included in the externals directory and is built automatically on Windows. On Linux or Mac, the expat library available in the system is used instead by default.

• Standard EM physics lists now require the G4EMLOW data files (version 6.23), in order to access precise cross-sections for Bremsstrahlung.

• All classes describing interactions of electrons, positrons and photons (Livermore and Penelope based) have been upgraded to the common software design adopted by both Standard and Low-Energy electromagnetic domains. Users are invited to use these upgraded classes following the migration procedure described in the following page: https://twiki.cern.ch/twiki/bin/view/Geant4/LoweMigration. The old obsolete classes have been removed from the Geant4 kernel. Users are invited to contact the Low Energy Electromagnetic working group in case of difficulties with this upgrade.

• A new version of the GDML schema (GDML-3.0.1) is included in this release. The GDML reader and writer are now compliant with GDML-3.0.1.

• Number of reference physics lists is reduced. Nearly all EM 'variant' physics lists are no longer available as a class in the library (e.g., FTFP_BERT corresponds to a class in the library, FTFP_BERT_EMV does not exist as a class). However, EM options are all easily available from updated G4PhysListFactory, allowing to use any EM option on top of a base hadronic physics configuration (e.g., it is possible to request "FTFP_BERT_EMV" by name).
• The implementation of physics-list constructors (builders) has been revised. Ordering numbers for processes are no longer used. Old method is still allowed but it is recommended for a custom user's physics list to be revised to be compatible with the new mechanism.

• Obsolete signature of G4Exception (using only one, text string, argument) has been removed. Use of a signature with four arguments, including severity, is now mandatory. For code that is portable between previous releases and this release, it is necessary now to use the signature:
  void G4Exception(const char* issure, const char* errorCode, G4ExceptionSeverity severity, const char* comments)
  Note: the enum G4ExceptionSeverity has been defined since Geant4 release 5.0 and its values are: FatalException, FatalErrorInArgument, EventMustBeAborted, RunMustBeAborted, and JustWarning.

• New data set versions: G4EMLOW.6.23, G4RadioactiveDecay.3.4, G4PhotonEvaporation.2.2, G4PII-1.3, G4NDL.4.0, G4NEUTRONXS-1.1.

4. New Developments and Capabilities

• It is now possible to define volumes with material in parallel geometry layers. The volumes created can complement or overlap the standard 'main' geometry for tracking. This new capability is called a "layered mass geometry". Zero, one or more additional layers for the mass geometry can be used. This is an enhancement of the existing functionality for parallel specialised geometries, which has been used for scoring, fast simulation, shower parameterisation and event biasing. The registered parallel geometries are scanned now for material. A side effect of this new capability is that all geometries and all regions in the each world defined are now scanned for their material.

• Applications with many materials can reduce their memory footprint by using the new Base Material capability. With it, materials with the same composition but different density will share entries in the tables of all electromagnetic processes. Tables with cross-sections, stopping powers and ranges are precomputed only for base materials. Any number of materials with different names and densities can be built from one base material. So, for example,it is now possible to define large number of "water" materials with slightly different densities. Only the base "G4_WATER" physics tables will be built at initialisation; these will be scaled at run time for all derived materials .

• A new tune of the Urban multiple-scattering model for electrons is provided. Is is named G4UrbanMscModel95. This features higher accuracy of electron scattering and stability of energy deposition when the step size is varied. The model is currently set as the default for electrons and positrons.
• A new model for Bremsstrahlung, G4SeltzerBergerModel, is now used by default at energies below 1 GeV. The model is based on tabulated cross-sections published by S.M. Seltzer and M.J. Berger. It provides agreement with the low energy Livermore and Penelope models below 10 MeV and the standard relativistic model at 1 GeV, improving the most over the previous 'standard' Bremsstrahlung model in the energy range from 1 keV to 1 GeV.
• New biasing options are provided which can be used both with standard and low-energy electromagnetic physics. They include cross-section biasing, forced interaction, and secondary particle biasing. The secondary particle biasing consists of Bremsstrahlung splitting and Russian Roulette. All these options can be activated per process, particle, energy and geometrical region via UI commands or C++ code.

• Interactions of antibaryons (including antihyperons) and light antinuclei with matter have been implemented in the Fritiof (FTF) model. This model is valid for incident antibaryon energies from 0 to 1 TeV, and for incident antinucleus momenta from 150 MeV/c/nucleon up to 2 GeV/c/nucleon. Corresponding antibaryon and antinuclear cross section classes have also been added. New processes were added to handle inelastic reactions of antideuterons, antitritons, anti3He and antialphas.
• Cascade models
  • Bertini
    • The Bertini cascade is interfaced to the pre-compound and de-excitation module, enabling improved generation of low energy secondaries. This option is fully benchmarked, and is targetted for applications sensitive to the spectra of low energy nuclear products.
    • An interface for re-scattering the fragments of the high energy string models has been created. This is an early stage development.
    • Bug fixes in the three-body final state reactions provided slightly better angular distributions of final state hadrons.
    • Memory churn (allocation and de-allocation of memory) was significantly reduced throughout the model.
    • A new nucleon-coalescence model is available, for improved production of energetic deuterons and light ions. It is not the default in existing configurations, due to the extra CPU cost.
    • Gamma-nucleon channels were added for future use, but are not activated.
  • Binary
    • The excitation energy used for re-scattering was improved.
    • The large energy non-conservation observed for hydrogen targets was fixed.
  • INCL/ABLA
    • The first version of INCL++, based on the INCL4.6 model, was released. It supports proton, neutron and pion projectiles up to 3 GeV and uses the native G4ExcitationHandler for nuclear de-excitation after the cascade.
    • The INCL light ion model now supports incident ions up to 18-Oxygen.
• Nuclear de-excitation models have been improved to allow G4PhotoEvaporation to be used for nuclei with Z > 100. Users may now include their own photo-evaporation data files. Several improvements were also made to the Fermi-Breakup model to improve behavior for light (A < 17) target nuclei.
• Several elastic scattering cross-sections, models and processes were extended to handle antimatter projectiles. These include G4WHadronElasticProcess, G4AntiNuclElastic and G4CHIPSElasticXS.
• A new module which handles quasi-elastic scattering has been created. It is extracted and now separate from the CHIPS code where it originated. It is used by all physics lists.
• Low energy neutrons interactions based on data-libraries:
  • The Geant4 Neutron Data Library (G4NDL) is now based on the ENDF VII database exclusively. Its new version, G4NDL 4.0, is required for use with the High Precision (HP) models with this release.
  • New translations of alternative neutrons database (ENDF, JENDL, JEFF, MENDL, CENDL), are being made available separately from IAEA. These can be used with the current HP neutron models, for studies of systematics.
  • Thermal scattering is enabled in the HP neutron models and the appropriate data files have been added to ENDF/B-VII, with the exception of para (ortho) H and D, liquid (solid) methane and benzene.
  • New low energy neutron models (LEND) which use the Livermore neutron database have been released. They use the GIDI interface to the database. All information on this database will be maintained and updated by Lawrence Livermore National Lab.
• Radioactive decay has been extended to handle forbidden beta decays and spontaneous fission. The corresponding data set has been updated to reflect this. The radioactive decay code was also corrected to allow atomic relaxation of de-excited nuclei produced during alpha and beta decay. Bug fixes in the internal conversion code now provide better agreement with deexcitation gamma spectra. Also, new code and messenger commands now allow directional biasing of decay daughters, including e-, e+, gamma, n and alpha.
• Improved muon-nuclear reactions are implemented in G4VDMuonNuclearModel. Here, the LEP models that used to handle the meson interactions are replaced by the Bertini Cascade. A new cross-section class, G4KokoulinMuonNuclearXS, was developed which splits out the cross-section code which was formerly hard-coded into the old muon-nuclear process.
• Energy and momentum checking at the process and model level can now be enabled by environment variable.

• Improvements to the Qt Graphical User Interface and Viewer. Involves installation of Qt from qt.nokia.com and selection with CMake - see CMake instructions.
• Users are encouraged to use the generic "/vis/open OGL" command, which selects the OpenGL viewer most appropriate to the installation.
• Some new /vis/ commands - see Detailed list of changes and fixes.

• The new G4SeltzerBergerModel bremsstrahlung model is used by all standard EM builders below 1 GeV.
• The new G4UrbanMscModel95 model is used by all EM builders except for G4EmStandardPhysics_option2 in which the old G4UrbanMscModel93 model remains. In the G4EmStandardPhysics_option2 builder, G4WentzelVIModel is used for multiple scattering of protons, pions and kaons. The Rayleigh scattering process is added to G4EmStandardPhysics_option3 builder.
• The physics-list factory allows to select one of EM options on top of base hadronic physics configuration. "_EMV" - standard option1, "_EMX" - standard option2, "_EMY" - standard option3, "_LIV" - Livermore, and "_PEN" - Penelope.

5. Expected effects on physics and computing performance

• Improved the accuracy of the Bremsstrahlung simulation and low-energy electron transport. Due to the adoption of the new model for Bremsstrahlung, the kernel part of the electromagnetic shower appears slightly narrower, no effect on the tails of the electromagnetic shower are observed.
• Retained similar CPU performance to the previous Geant4 versions.

• The recommended physics lists for HEP applications (FTFP_BERT, QGSP_FTFP_BERT and QGSP_BERT), show results very similar to the previous production release. A small decrease in energy deposited in calorimeters is expected for FTF-based physics lists with respect to the previous release for primaries with energy above 20 GeV. The decrease is of order ~1%.

6. Known Run-Time Problems and Limitations

7. Compilation Warnings

8. Known Run-Time Warnings

    G4Transportation: Statistics for looping particles 
       Sum of energy of loopers killed: 
       Max energy of loopers killed: 

9. Geant4 Software License

10. Detailed list of changes and fixes

• CMake:
  • Set default to build shared libraries on all platforms.
  • Added support for Xm Motif and OpenInventor drivers.
  • Updated scripts to export use of internal CLHEP when used. Option to use system CLHEP is automatically set to ON if the user has supplied a CLHEP_ROOT_DIR argument.
  • Added script UseGeant4.cmake providing a simple CMake module which users can include to setup the build of a basic Geant4 application. It sets definitions, include directories and compiler flags.
  • Added support for generation of DLL libraries on WIN32 platform.
  • Abilitated built-in dylib symbols on Mac for shared libraries; it is now no longer necessary to specify paths in DYLD_LIBRARY_PATH.
  • Automatically use GDML if user supplies XERCESC_ROOT_DIR argument as path to XercesC installation. Added support for setting GDML/Xerces variables as needed.
  • Providing full setup of environment from build and install trees, allowing developers to build applications directly using the build tree.
  • Fix for Unix platforms in setting the output directory for libraries. Partially addressing problem report #1200.
  • Added universal shell script skeleton for generating backward compatible environment variable setup. Addressing problem report #1202.
  • Added 'uninstall' target to remove all items installed via the install command.
  • Removed unsetting of data library environment variables in generated environment scripts.
  • First version of CPack configuration for binaries. Added TBZ2 and ZIP types back into list of source package generators.
  • Added initial set of files needed for building tests and examples with CTest.
  • Added the downloading and installation of Geant4 Data files if the option GEANT4_INSTALL_DATA is set. Added guard on option and install of data so that this is only available on CMake >= 2.8 due to use of ExternalProject.
  • Forced all CXX flags to use those used to build Geant4. Addressing problem report #1227.
  • Corrected setting of INSTALL_NAME_DIR for library targets. Addressing problem report #1231.
  • Added build of new "analysis" module.
  • Added protection for setting of G4LIB_BUILD_DLL to happen on WIN32 only.
  • Added support for tests on Windows with NMake and Visual Studio.
  • Added Geant4CTest.cmake module for the configuration of CTest and added pre- and post-commands for runnig test/examples.
  • Added possibility to generate Root dictionaries in FindROOT.cmake macro.
  • Moved setting of initial values for TestRelease and Maintainer mode compiler flags to make rules override file. Values are now set based on compiler and platform. Added TestReleaseDebug build mode.
  • Removed enabling of Fortran language when G3toG4 is enabled.
  • Added G4VIS_USE and G4UI_USE to list of core compile definitions conditional on the existence of the variables G4VIS_NONE and G4UI_NONE to match original GNUMake behavior.
  • Added 'ui_all' and 'vis_all' components to allow simple selection of all available UI and Vis components.
  • Added support for OpenInventor visualization driver, with new CMake option GEANT4_USE_INVENTOR to enable build. Addressing problem reports #1196, #1261.
  • Added setup of installed scripts to use self location functionality in geant4-config when installation is relocatable.
  • Added a function to conditional set environment variables. Added setup of G4WORKDIR. Added setup of data library variables if they've been installed. Added GEANT4_ADD_FEATURE reporting for data and example code install.
  • Marked GEANT4_ENABLE_TESTING and GEANT4_BUILD_EXAMPLES options as advanced.
  • Updated references to data libraries version for 9.5.
• GNUMake:
  • Added default for CLHEP_BASE_DIR to built-in package in externals. Integrated flags in architecture.gmk and binmake.gmk for use of the CLHEP built-in package, and adapted setup for build of DLLs on Windows. Added flags to handle future driver OpenGLWt.
  • architecture.gmk: corrected definition of G4LIB_USE_GDML, to properly activate GDML settings.
  • Darwin-g++.gmk: modified OGLLIBS settings to support MacOSX 10.7 Lion for proper configuration of OpenGL libraries in combination with Qt. Allow use of Qt libraries not installed as Framework
  • Updated binmake.gmk to add new "electromagnetic/dna", new "hadronic/models/quasi_elastic" and "analysis" modules. Changed common.gmk to take into account paths to headers.
  • G4UI_BUILD.gmk: added setting of G4INTY_BUILD_XT for OPENGLXM and OIX. Visualization drivers need Xt, so this covers the case when the user does not build Xt from G4UI_BUILD flags. Addressing problem report #1250.
  • Abilited possibility to force building of internal 'expat' module also for Linux and Mac, through use of G4LIB_BUILD_EXPAT flag.
  • Archived Configure script.

• New module providing managers for generation of histograms in the form of Root n-tuple/trees and AIDA XML, based on tools from the Inlib/Exlib package.

• Moved copy-ctor and operator=() for G4VReadOutGeometry to protected, as not meant to be directly used.
• G4THitsMap: removed the print-out of total value, to avoid inconsistency caused by a class which does not have += operator or cannot be initilized to 0. Addressing problem report #1258.
• Fixed Coverity defects: mainly uninitialised static data.
• Migrated code to new G4Exception scheme.

• Fixed compilation warnings on gcc-4.6 for unused variables.
• Fixed defects reported by Coverity.
• Migrated code to new G4Exception scheme.

• Added method GetTranfMat() in G4ErrorFreeTrajectoryState. Addresses problem report #1248.
• Fixed Coverity defects for uninitialised static data and cases of ios format not restored.
• Migrated code to new G4Exception scheme.

• Modified G4ParticleGun, G4PrimaryTransformer, G4RayShooter and G4SingleParticleSource to use kinetic energy and momentum direction.
• Fixed minor bugs in G4HEPEvtInterface: in creation of G4PrimaryParticle, only PDG code is given to the constructor; then mass and momentum are set.
• Some code cleanup in G4ParticleGun and G4SPSPosDistribution.
• G4ParticleGun: added const qualifier to accessors.
• Fixed Coverity defects for uninitialised static data. Fixed compilation warnings on gcc-4.6.
• Migrated code to new G4Exception scheme.

• New module for optional external packages. Currently including:
  • Built-in "CLHEP" module, including a subset of CLHEP classes imported from version CLHEP-2.1.1.0 and adapted to the Geant4 build scheme.
  • Built-in "expat" module, including source code from original 2.0.1 version of expat. Required only on WIN32 for "lend" hadronic model.

• Added export flag for extern symbols to allow for generation of DLLs on Windows; allow build of DLL, if selected. Fixed signature in PG4gsdet() for unresolved symbol in building DLLs.
• Make safe to compile with -j and drop compilation of Fortran code.
• Use canonic indexed form for G4Exception.
• Fixed Coverity defects.

• Migrated to new G4Exception scheme.
• Fixed Coverity defects.
• Biasing
  • G4ProcessPlacer: using SetProcessOrderingToSecond() for biasing process, and leave transportation as first.
• Cuts
  • Introducing parallel layered mass geometry. Material-cuts-couple is now updated for all worlds.
• Decay
  • Added sub-type for radioactive decay 'DECAY_Radioactive' (equal to 210).
  • Fix in G4Decay to propose the final local time rather than the global time so that local time can be updated properly. Addressing problem report #1224.
• management
  • Suppressed warning message of G4ProcessTable::Insert() in case of adding ions with inactivate processes. Addresses problem report #1173.
  • Added check for duplicate call for SetOrderingToFirst(Last).
  • Fixed memory leak in copy constructor of G4ProcessAttribute.
• Optical
  • A dielectric_dielectric surface may now have a wavelength dependent property "TRANSMITTANCE". If this is specified for a surface it overwrites the Snell's law's probability. This allows the simulation of anti-reflective coatings.
  • Added logic for ProposeVelocity() to G4OpBoundaryProcess::PostStepDoIt().
• Parameterisations
  • Fixed bug in G4FlashShowerModel::ModelTrigger() for energy condition.
• Scoring
  • Introducing parallel layered mass geometry. G4ParallelWorldProcess takes care of both layered mass geometry and sensitivity of a parallel world.
• Transportation
  • Defined ProcessSubType for transportation processes.
  • Changed process type of G4NeutronBuilder to fGeneral.
  • Change in G4Transportation and G4CoupledTransportation to propose the final local time rather than the global time so that local time can be properly updated. Addressing problem report #1224.
  • Modified calcuation of deltaTime for optical photons in G4Transportation (and G4CoupledTtansportation). The final velocity obtained by G4Track CalculateVelocityForOpticalPhoton() should be used instead of the initial velocity of the step. Addressing problem report #1239.
  • G4Transportation and G4CoupledTransportation: allow for spin tracking of neutral particles with a magnetic moment, turned off by default, can be enabled by using the new method EnableUseMagneticMoment(). Allow for uniform gravity field tracking of neutral particles. Removed all special treatment and logic for opticalphoton.

• Migrated to new scheme of G4Exception for warnings and errors.
• Divisions
  • Optimised initialization of phi angles in parameterisation for tubs and cons.
• Magnetic field
  • Added capability for (uniform) gravity field. Implemented with new classes G4EqGravityField, G4UniformGravityField.
  • G4Field: revised to help detect the presence of a gravity field. Added new method IsGravityActive().
  • Corrected computation of omegac in G4EqEMFieldWithEDM, G4Mag_SpinEqRhs and G4EqEMFieldWithSpin with more coherent formula. Simplified formula for computation of kinetic energy in G4FieldTrack.
  • Avoid printing out pointer in G4CachedMagneticField::ReportStatistics().
  • Added new class G4TrialsCounter for keeping statistics of step trials.
• Management
  • Introducing parallel layered mass geometry. Scanning of material is extended to all the regions of all the worlds.
• Navigation
  • Fix in locator classes for tracks stuck on boundaries between volumes: improved the condition for accepting a candidate intersection with a boundary in all locator classes. It now checks whether the track enters in the new volume, using GetGlobalSurfaceNormal() from G4Navigator.
  • Revised G4VIntersectionLocator: added new method GetSurfaceNormal(), to get normal at last candidate boundary crossing. Corrected AdjustmentOfFoundIntersection() to use exit-normal in global coordinates instead of local.
  • Modified G4Navigator::GetLocalExitNormal() method, to allow it to be called before one of the Locate...() methods is called ( LocateGlobalPointAndSetup() or LocateWithinVolume() is called. Returns the normal at the endpoint of the last ComputeStep(); Added new method to provide the exit-normal in global coordinates; added new methods GetMotherToDaughterTransform() and GetLocalExitNormalAndCheck().
  • Adapted G4MultiNavigator to revisions of intersection logic to provide exit-surface normal for step ending on boundary.
  • Trivial fixes for Coverity defects in G4GeomTestPoint and G4GeomTestVolPoint.
  • Some code cleanup.
• Solids (Boolean)
  • Corrected protection in G4SubtractionSolid::DistanceToIn(p,v). Revised outcome, to not cause event abortion, in case of bad convergence.
• Solids (CSG)
  • First implementation of G4CutTubs solid, a tube with possible cuts in +-Z.
  • G4Sphere: fix for setting of relative radial tolerances when setters for radiuses are explicitly called (affecting parameterisations on radius for spheres). Relaxed validity condition in construction for the minimum size of the outer radius.
• Solids (Specific)
  • Fix in G4TriangularFacet::Intersect() to take into account geometrical tolerance and cases of zero distance from surface's facet; fixed tests where square root was compared to half Cartesian tolerance; now compared to kCarTolerance.
  • Fix in G4TessellatedSolid::DistanceToOut(p) and DistanceToIn(p) to compute exactly the distance from facet avoiding use of shortcut/variant using the 'outgoing' flag. Addresses problem report #1242.
  • Fix in G4Paraboloid::DistanceToIn(p,v) and DistanceToIn(p) to take into account tolerance correctly and provide better estimation of distance. Addresses problem report #1234.
  • Added SetReferences() and InvertFacets() to G4GenericTrap::CreatePolyhedron() for correct visualisation of Boolean compositions with G4GenericTrap.
  • Added SetReferences() to G4TessellatedSolid::CreatePolyhedron() for correct visualisation of Boolean compositions with G4TessellatedSolid. Addresses problem report #1235.
  • Fixed logic in G4VTwistSurface::GetEdgeVisibility() for the treatment of corners.
• Volumes
  • Use WIN32 instead of G4LIB_BUILD_DLL as flag for protecting allocator choice in G4NavigationHistory on Windows.

• Added banner and footer for errors and warnings, to allow automated parsing for the start and end of all printout for all calls to G4Exception. Redirected warnings to cout instead of cerr by default.
• Removed obsolete G4Exception signature, that took simple string as argument. Added two new G4Exception signatures, which take a ostringstream for any amount of additional output.
• G4Physics2DVector, G4Physics2DVectorCache: new classes implementing a 2-dimensional vector with linear interpolation.
• Added dBin, baseBin and verboseLevel in the base class G4PhysicsVector to avoid having additional properties data introduced in derived classes. This enables use of G4Allocator for allocating G4PhysicsVector and G4PhysicsVectorCache objects. Fixed ScaleVector() method in G4PhysicsVector derived classes.
• G4PhysicsLogVector, G4PhysicsLnVector, G4PhysicsLinearVector: updated to safer implementation of ScaleVector() method.
• Added protection in G4Pow::powN() method for high exponent values. Reduced vector for factorial from 512 to 170 (result should be below DBL_MAX). Fixed computation of log(factorial).
• Added G4lrint(), G4lint(), G4rint() as global utility functions for generic use.
• Added missing post-const qualifier to G4String::compareTo() methods.
• Update date of release for 9.5.

• G4PolyhedronArbitrary: added SetReferences() and InvertFacets() methods. SetReferences() is necessary at the end to complete the polyhedron, in particular if the polyhedron suffers subsequent Boolean operations.
• G4Polymarker, G4VMarker: added copy-constructors.
• G4Colour: added GetMap() so that current colour map options can be listed.
• Migrated to new scheme of G4Exception.
• Fixed most Coverity warnings.

• Completed migration to integer Z and A.
• Added ModelDescription() method to all model classes for HTML documentation.
• Migrated to new scheme of G4Exception.
• Fixed compilation warnings on gcc-4.6 and Coverity reports.
• Cross sections
  • Implemented re-design of G4VCrossSectionDataSet interface.
  • New class G4KokoulinMuonNuclearXS: derived from code in the direct-implementation process G4MuNuclearInteraction, replacing the old-style implementation for conforming to the hadronic framework design of process, model, cross-section. This new class is now the default cross-section for G4MuonNuclearProcess.
  • Provide descriptions for nearly all cross-section classes in method CrossSectionDescription().
  • G4CrossSectionElastic, G4CrossSectionInelastic and G4ComponentAntiNuclNuclearXS: adopt modified names and fixed usage in physics-list. Fixed problem of double-deletion.
  • G4VComponentCrossSection: changed method names to GetTotalElementCrossSection() and GetTotalIsotopeCrossSection().
  • G4IonsShenCrossSection: extended energy range up to 100 TeV.
  • G4IonProtonCrossSection: implemented all virtual IfApplicable() methods.
  • Added GetName() method in G4VCrossSectionDataSet, to be used with documentation reporting.
  • G4CrossSectionDataStore: fixed selection of isotope when element cross-section is used but isotope composition is defined by the user. Fix in DumpPhysicsTable() method, responsible for incorrect ion cross-sections values when invocated.
  • G4IonsShenCrossSection added method GetIsoCrossSection().
  • Modified p-p cross sections in G4GGNuclNuclCrossSection, G4GlauberGribovCrossSection and G4HadronNucleonXsc to include latest TOTEM data at 7 TeV.
  • G4ElectroNuclearCrossSection: corrected numerical instability adding numerical protections in methods computing equivalent photon energy, Q^2, and virtual factor.
  • Extended G4GlauberGribovCrossSection and G4HadronNucleonXsc for kaons.
  • G4ComponentAntiNuclNuclearXS: introduced Plab/baryon. Fixed anti-baryon-nucleon cross-sections at low energies at value they have at 100 MeV/c of projectile/anti-nucleon.
  • G4UPiNuclearCrossSection: fixed Barashenkov data for pi+.
  • G4BGGNucleonElasticXS, G4BGGNucleonInelasticXS: fixed cross-section off Hydrogen and pi+ cross-sections; use PDG and GHEISHA parameterisations.
  • G4TripathiLightCrossSection: added protections against big positive values of cross-section below Coulomb barrier.
• Management
  • G4HadronicProcess: make GetCrossSectionDataStore() public so it can be accessed through the process. Added method MultiplyCrossSectionBy() so that cross-sections can be scaled for testing.
  • G4HadronicProcess: added public method GetTargetNucleus() to access target isotope; added private method FillResult() providing simpler and more transparent handle for primary and secondaries. Added null-pointer checks for material and volume to DumpState().
  • G4HadronicProcess: completed migration to integer Z and A. Added file-scope variable to capture envvar with name of file for saving random engine state. Used immediately prior to model ApplyYourself() call.
  • G4HadronicProcess: changed order of checks; first track status, then value of cross-section to remove potential numerical problem of cross-section computation for zero kinetic energy hadrons. Added change status if final energy of primary particle is zero Added protected method GetTargetNucleusPointer().
  • Implemented migration to new G4HadFinalState interface for G4HadronicProcess and related classes. Adopt new interface to cross sections; use GetElementCrossSection() in place of GetMicroscopicCrossSection(); added check on zero kinetic energy of primary and secondary after interaction.
  • Allow to turn on checking for E/p conservation by environment variables. Take the tighter limit of process and model. Enhanced printout for case 3 in E/p checking with initial state info.
  • G4HadronicProcessType: added 'RadioactiveDecay' with ID 210 to enumeration.
  • Extended Print() method in G4HadronicProcessStore to include cross-section info. Added method DumpHtml() to write out html physics list documentation. Added also method PrintHtml() to print html documentation for processes, models and cross sections for each particle.
  • Models:
    • E/p checking extended to allow string models to set limits; GetEnergyMomentumCheckLevels() made virtual in G4HadronicInteraction, and implemented to query string model in G4VHighEnergyGenerator.
    • Added data member in G4VIntraNuclearTransportModel with related accessor methods to store/retrieve original projectile.
    • G4VIntraNuclearTransportModel: fixed initialisation of primary projectile.
    • Added constructor and virtual destructor to G4VIsotopeProduction.
• Processes
  • Added G4MuonNuclearProcess, meant to replace the old process G4MuNuclearInteraction by separating model and cross-section classes.
  • Added processes for light anti-ions: G4AntiAlphaInelasticProcess, G4AntiDeuteronInelasticProcess, G4AntiHe3InelasticProcess and G4AntiTritonInelasticProcess.
  • G4WHadronElasticProcess: moved from coherent_elastic; implemented random rotation of final state. Added change status if final energy of primary particle is zero. More accurate handling of low-energy particles.
• Utility
  • G4ReactionProduct: use G4Allocator for dynamic allocation to reduce memory churn.
  • G4HadFinalState: store vector of values instead of pointers to reduce memory churn. Improved constness of argument passing and functions, and reduced creation of temporary objects. Drop unneeded staleness flag. Removed AddSecondary() interface. Added interfaces to merge two final states together.
  • Models:
    • New utility class G4DecayKineticTracks, to be used by models' Propagate() member function, scanning input list and decaying any short-lived resonances in situ. Parent particles are deleted from the list at following decay. Adapted G4DecayStrongResonances accordingly.
    • G4Fancy3DNucleus: reduced memory churn by eliminating duplicate arrays of G4Nucleon, moving the Helper buffer and other buffers into class data members.
• Binary Cascade
  • Set E/p checking limits for energy non-conservation checks. Fixed large energy non-conservation for Hydrogen target (vector for the final state was not cleared before use). Also moved check for Hydrogen to ApplyYourSelf().
  • G4GeneratorPrecompoundInterface: added code to calculate excitation energy as difference of initial-final state, to allow for energy conservation for string models. Reduced problem of energy non-conservation. Fix to add fragment nucleus to reaction products when no de-excitation. Replaced loop to decay input secondaries with new G4DecayKineticTracks utility. Fixed memory leak in to correctly delete kinetic track.
  • G4BinaryCascade: revised calculation of excitation energy when used for rescattering, removing systematic energy non-conservation. More changes/fixes for use with propagate interface. Added handling of detroyed nucleus, i.e. when charge becomes 0. Some code cleanup, mostly ApplyCollision().
  • Fixed bug in G4RKFieldIntegrator, following Xcode warning.
• Bertini Cascade
  • Developed infrastructure to support using Bertini as a "rescattering" model for the output of the FTF simulation.
  • Make cross-section tables self-registering at job initialization.
  • G4NucleiModel: added optional argument to reset() with list of collision points, for copying cascade state when rescattering. Optimised computation of "rho". Reverted length scales to "2.82" factor. Compared local kaon/hyperon cross-section tables with sum-of-exclusive tables from G4Cascade*Channel classes. Fixed bug in G4InuclNuclei constructor (mom,A,Z,Eexc,model) to properly set excitation energy; caused energy non-conservation in de-excitation code. Changed kaon potential to be positive, not negative, so slow kaons will occasionally be captured. Removed obsolete local cross-section arrays; use standard tables, which are numerically equivalent.
  • G4CascadeInterface: added support for capturing random-engine state for debug at each ApplyYourself() or Propagate() invocation. Simplified NoInteraction() to have no particles returned, but use SetEnergyChange() to communicate to tracking that real track should be stopped and deposit all its energy. Only projectile is copied to final state, not target nucleus. If accept/reject loop fails (maximum retries), return initial state untouched rather than last generated cascade. This avoids exposing an elastic scattering event under any circumstances. Use new G4DecayKineticTracks in Propagate() to handle short-lived inputs. Set up for E/p checking by G4HadronicProcess in constructor; set internal E/p limits (currently 5% and 10 MeV).
  • Simplified cascade test-and-repeat loops; do four-momentum balancing with nuclear fragment, now used exclusively as G4Fragment, at end of cascade, not after de-excitation.
  • G4CascadeCheckBalance: added non-limit ctor and explicit limit setting.
  • G4IntraNucleiCascader: added list of hit-nucleon locations, filled by copyWoundedNucleus() for initialization of G4NucleiModel. Reduced number of retries from 1000 to 100. For pre-cascade rescattering, put resonances, or any particle which cannot be propagated by Bertini, directly onto output, without decaying.
  • G4CollisionOutput: added non-const accessors for lists.
  • G4LorentzConvertor: added diagnostic messages in rotate(). Drop most "intermediate" kinematic parameters used in rotate, in favor of somewhat more direct vector operations.
  • G4CascadeChannelTables: use FindTable() to determine if table has been created already or not. Added new gamma-nucleon interaction states.
  • G4CascadeGamNChannel, G4CascadeGamPChannel: new final state tables for gamma-nucleon interactions, with final states up to 9-body, using the same energy binning as pion-nucleon.
  • G4Cascade*Channel: added new constructor argument for all cross-section tables. Removed all namespace self-instantiations to fix incosistencies in initialization across platforms.
  • G4Inucl*Particle: modified all constructors to take the model code as an optional argument, so that subclasses don't have to set it manually.
  • Fix in G4ElementaryParticleCollider to hide no longer used "mult" argument to getMomModule2toMany(); removed special "mult==3" case.
  • Fix in G4CascadeT1NNChannel to final 9-body final-state list for both nn and pp: existing entries violate charge.
  • Added protections in G4LorentzConvertor for rotate() applied to at-rest system (i.e., z-axis ~ (0,0,0)), and simplified calculations.
  • Updated to reduce memory churn in G4IntraNucleiCascader, G4FissionStore, G4Fissioner, G4CascadeInterface, G4InuclElementaryParticle, G4BigBanger, G4ElementaryParticleCollider.
  • G4EquilibriumEvaporator: improved building of final-state particles to reduce creation of temporaries.
  • G4CollisionOutput: use resize() and set/fill to eliminate temporary copies. Added diagnostics and fixed bug with momentum units.
  • G4CascadParticle: added functions to explicitly set all data members, including initial path and generation, to support changes below. Re-order member functions in set/get pairs. In getPathToTheNextZone(), check for zero momentum (pp < 1e-9) and return zero path-length; this will support capture at rest.
  • G4CascadeCoalescence: NEW factory class to implement conversion of final-state nucleon "clusters" into light ions. Uses LAQGSM algorithm (Sec. 2.3).
  • G4ElementaryParticleCollider: added protection against non-physical arguments and FPE in two-body angular distributions. Moved message about unrecognized initial-state behind verbosity flag.
  • G4IntraNucleiCascader: added final-state clustering as optional facility, controlled through compile-time flag G4CASCADE_DO_COALESCENCE. Clustering is done during cascade finalization.
• Chiral Invariant Phase Space (CHIPS)
  • Tuning of the Quasmon/Evaporation transition.
  • Correction for Coulomb Barrier. Addressing problem report #1192.
  • Implemented migration to new G4HadFinalState interface in G4ElectroNuclearReaction. Fix in destructor.
  • G4QHadron(Elastic/Inelastic)DataSet: returning '0' for missing hadrons.
  • Fix in G4QEnvironment for object deletions. Fixed initialization in constructor.
  • Fixed case of FPE for unprotected sqrt() with possible negative argument in G4QNucleus. Addressing problem report #1251.
  • Corrected G4QKaonPlusElasticCrossSection and G4QPionMinusNuclearCrossSection for a right reproduction of K+N and Pi-N cross sections.
  • Correction in G4QNeutronNuclearCrossSection, G4QProtonNuclearCrossSection for right reproduction of cross-sections, as in G4QuasiElasticRatios.
• De-excitation
  • G4NeutronRadCapture: fixed in particle change, kill primary particle and save residual nucleus.
  • G4ExcitationHandler: fix for Fermi break-up products de-excitation. Added protection for small excitation energy into photon evaporation loop. Updated logic: first call only to MFM model, second start evaporation loop, where FBU is called for light fragments, products of FBU are sent to the 3d loop on photon evaporation.
  • General cleanup of the FermiBreakUp model: Removed G4FermiSplitter and G4FermiIntegerPartition. Modified G4FermiConfiguration to be a simple container. Removed unused methods in G4FermiConfigurationList and fixed minor memory leak. Removed map and access to map methods in G4FermiFragmentsPool. Extended printout when kinematics fail in G4FermiPhaseSpaceDecay; fixed computation of majorant in BetaKopylov method, general cleanup.
  • G4VGammaDeexcitation: added accessors for the ICM flag, commented out usage of this flag during sampling.
  • G4Evaporation: directly send low Z,A fragments to the FermiBreakUp model. Added extra protection: do not compute GEM probability for the case Eexc > 3MeV*A; fixing FPE problem for rare FTFP events.
  • Fixed numerical problem for exp() function in G4EvaporationProbability.
  • G4NuclearLevelStore (photo-evaporation): added method to include user's evaporation data file; the given data will be used for the treatment of the photo evaporation of the given isotope instead of using the file provided in the photo-evaporation database.
  • Extended use of G4PhotoEvaporation to Z>100. Note that for Z>100 at the moment the binding energy of the vacant e- shell left after internal conversion is set as 0. and the atomic relaxation will not be simulated later on.
• Elastic scattering
  • G4WHadronElasticProcess: added light anti-nuclei; using particle names in the IfApplicable() method. Added NIEL energy deposition if recoil is below the threshold. Follow migration of G4HadFinalState interface. Set lowestNeutronEnergy to 1.e-12*eV. Added protection for very low primary energy after scattering. Moved class to 'processes' sub-module.
  • New class G4AntiNuclElastic for simulation of elastic anti-A - A scattering.
  • G4HadronElastic: do not create an ion if the recoil energy is below the threshold; fixed SampleInvariantT() default implementation to be used in testing. Lowest limit is set to 1.e-6 eV.
  • G4NuclNuclDiffuseElastic, G4DiffuseElastic: updated to use G4HadronElastic as the base class and to follow its interface.
  • Removed obsolete class G4UHadronElasticProcess.
  Renamed G4VHadronElastic to G4HadronElastic (main interface).
  • G4CHIPSElasticXS: added calculations of anti-baryons, Pi+, Pi-, K+, K- nucleus elastic cross-sections. Use G4HadronicException; added cross-sections for neutral kaons.
  • G4ChargeExchangeProcess, G4CHIPSElasticXS: implemented migration to new design of G4VCrossSectionDataSet interface.
  • G4LEpp, G4LEnp: fixed extraction of A from G4Nucleus().
• High Energy Theoretical
  • Store primary particle, so that cascade in Propagate() has access.
  • Added GetEnergyMomentumCheckLevel() method to G4TheoFSGenerator.
  • G4QuasiElasticChannel, G4TheoFSGenerator: reduced external dependences. Modified according to G4V3DNucleus::GetNucleons() signature.
  • G4QuasiElasticChannel: use QuasiElRatios from 'quasi_elastic' model in place of QuasiFreeRatios from CHIPS.
• INCL/ABLA
  • Use inverse kinematics in INCL light ion collisions: it improves INCL light ion model performance in applications where we are interested in beam fragmentation.
  • INCL light ion model now supports light ion projectiles up to and including Oxygen-18.
  • Bug fixes in ABLA for rounding problem, causing evaporation function to modify nucleus excitation energy via a side-effect.
• INCLXX
  • First version of INCL++ intra-nuclear cascade model based on INCL4.6 cascade model. Supported projectiles: protons, neutrons, pions up to 3 GeV. Uses G4ExcitationHandler for de-excitation.
• Isotope Production
  • G4NeutronIsoIsoCrossSections: implemented migration to new design of G4VCrossSectionDataSet interface.
• Lend
  • First implementation of LEND (Low Energy Nuclear Data) cross-section and physics model.
• lll_fission
  • Enabled to handle excited isomer nuclei as targets.
• Low Energy
  • Added protection against return back only primary particle with increased energy in G4LEDeuteronInelastic, G4LETritonInelastic and G4LEAlphaInelastic.
• Neutron High Precision
  • Added reaction Q value and break up flag (MF3::QI and LR) in G4NeutronHPInelasticCompFS.
  • Enabled FSMF6 in capture final state (G4NeutronHPCaptureFS).
  • Bug fix in G4NeutronHPFinalState for unexpected baryon balance. Fix for setting an index of the last element of an array.
  • Bug fix in G4NeutronHPPhotonDist for call to element with negative index.
  • Modified G4NeutronHPIsoData, G4NeutronHPFission and G4NeutronHPorLFission for ENDF-VII.
  • Relaxed checking condition of inconsistency in G4NeutronHPElasticFS, to avoid rounding problems on 32-bits systems. Added protection to avoid producing 0 kinetic energy neutrons.
  • Added protection for very low "remaining_energy" and fixes in G4NeutronHPContAngularPar.
  • Fixes in G4NeutronHPInelasticCompFS and use QI value for calculating excitation energy of residual.
  • Fixes for migration to ENDF-VII.r0. Bug fix in G4NeutronHPPartial.
  • G4NeutronHPFissionFS: modified according to new G4HadFinalState interface.
  • Migrated to new G4VCrossSectionDataSet interface.
  • Added cache for kinetic energy, element, material and cross-section.
  • Enabled to handle excited isomer nuclei as targets.
  • Migrated G4NeutronHPFastLegendre to high order (>30) Legendre polynominal.
  • Enabled to use Thermal Scattering Model and Data with NIST Materials.
  • Added ENDF/B-VII thermal scattering materials; excluding para (ortho) H and D, liquid (solid) methane and benzene.
  • Fixed FPE error in G4NeutronHPThermalScattering and memory leak in G4NeutronHPThermalScatteringData.
  • Requires new data set G4NDL-4.0.
• Parton-String
  • First implementation of barion - anti-barion annihilation. Added new G4FTFAnnihilation class in the FTF model, and made corresponding changes in related FTF classes.
  • Implemented simulation of anti-barion - nucleus interactions from the rest to 1000 GeV. Reggeon cascading is switched off. Good results are obtained in combination with the Binary cascade model.
  • Implemented first variant of FTF code for simulation of anti-nucleus-nucleus and nucleus-nucleus interactions. The anti-nucleus nucleus part is working starting from 150*A MeV/c. The nucleus-nucleus part is applicable at Plab > 2 * A GeV/c.
  • Decreased low energy limit for FTF.
  • Small improvement introduced in G4FTFParameters in order to treat Labmda-Bar, Sigma-Bar, etc...
  • Tuned parameters of Kaon-Proton interactions in FTF. CHIPS cross sections are used now in FTF for needed X's; created corresponding class: G4ComponentCHIPShadronNuclearXS.
  • Set limits for E/p checking.
  • G4QGSParticipants: reduced energy of participant nucleons by binding energy. This increases excitation energy, as calculated by initial-final. Without this, excitation E often is close to 0, or negative.
  • G4VParticipants: removed external dependences on G4Fancy3DNucleus.
  • G4GammaParticipants: updated according to new G4V3DNucleus::GetNucleons() signature..
  • G4LundStringFragmentation: made public and factorized the virtual method FragmentString(). Factorized method SplitLast() to improve maintenance. Simplified GetLightConeZ() implementation.
• Photo-lepton-hadron
  • New model G4VDMuonNuclearModel to replace old-style G4MuNuclearInteraction and G4ParametrizedHadronicVertex based on GHEISHA.
  • Removed G4MuonNucleusProcess and G4MuonNucleusInteractionModel classes. These were the old GHEISHA-style models no longer used.
• Pre-equilibrium
  • G4PreCompoundModel: added limit on nucleon number maxZ=3, maxA=5, allowing pre-compound emission for light fragments (old formulation of the model). Restored soft-cutoff option and use "go_ahead". Set low limit for excitation to 1 MeV.
  • G4PreCompoundTransitions: return zero probability if no excitons.
• QMD Reaction
  • Migrated to new G4VCrossSectionDataSet interface.
• Quasi Elastic
  • New module including the quasi-elastic model extracted from CHIPS. It is used by all physics-lists and it now made independent from CHIPS.
• Radioactive Decay
  • G4RadioactiveDecay: added data members and Set/Get functions to support directional biasing ("collimation") of visible decay daughters (e+, e-, gamma, neutron, and alphas, but not neutrinos or the residual nucleus). Added corresponding commands in G4RadiactiveDecaymessenger. Set process subtype (fRadioactiveDecay = 210). Fix for local decay time; addressing problem report #1224. Added new Fermi function and forbidden beta decay spectral shapes. Added possibility for the user to load its own decay file. Set halflifethreshold negative by default to allow the tracking of all excited nuclei resulting from a radioactive decay. Added reading of 'Spontaneous Fission' mode from data files. requires new data set G4RadioactiveDecay.3.4.
  • G4RadioactiveDecayMode: added SpFission mode.
  • Created G4BetaDecayType, an enum of beta transition types.
  • Created G4BetaDecayCorrections, a class containing new Fermi function and forbidden transition calculations.
  • Added check for maximum shell number in G4NuclearDecayChannel.
  • G4RadioactiveDecayChannel: correction to allow the atomic relaxation of de-excited nuclei produce also during beta and alpha decay.
  • G4RadioactiveDecayMessenger: added macro commands for setting files.
  • G4NuclearDecayChannel: migration to the new design of atomic deexcitation, activation of deexcitation via EM UI commands or G4EmProcessOptions as it is shown in EM examples. Added check if de-excitation is enabled.
  • G4NucleusLimits: suppressed Z<100 limit.
• Stopping
  • Added class G4FTFCaptureAtRest to handle nuclear capture of anti-protons at rest with Fritiof.
  • Added G4PiMinusAbsorptionBertini class.
• Utils
  • Added new methods in G4Nucleon for creation of anti-nucleus. They can be used to replace protons and neutrons into anti-protons and anti-neutron, respectively.
  • Moved G4lrint(), G4lint(), and G4rint() global functions to 'global' module for generic use.

• Migrated code to new G4Exception scheme.

• Added a flag in G4Qt to know if there is an external Qapp started.
• Fix for displaying a QMessageBox on G4err output, to allow for proper detection of G4Exceptions coming from kernel.
• Code quality improvements by Coverity.
• Migrated code to new G4Exception scheme.

• G4Material: added new "Base-material" approach allowing to share internal tables for dedx, ranges, cross section between similar materials; a pointer to the base material (NULL by default) and corresponding methods have been added; added also extra protections against incorrect mass fraction during material construction.
• G4NistMaterial: added constructor of a material using base material; reduced size of internal vectors. Cleanup in methods to build materials in G4NistMaterialBuilder.
• G4IonisParamMat: use mean ionisation potential provided by user in computation of density effect.
• Added new class G4ElementData, data structure for cross-sections, shell and isotope cross-sections.
• Reviewed and updated body tissue compositions according to ICRU Report 46 (1992) instead of old data (1975); renamed G4_TESTES_ICRP to G4_TESTIS_ICRP; removed G4_CYTOSINE, G4_THYMINE, G4_URACIL, as already existing as G4_DNA materials.
• G4Element, G4Material: reviewed comments. Addressing problem report #1220.
• G4Isotope: added flag 'fm' to identify isomer level.
• Removed G4MaterialPropertyVector implementation and make it a typedef of G4PhysicsOrderedFreeVector; removed G4MPVEntry class and modified G4MaterialPropertiesTable accordingly.
• G4SandiaTable, G4IonisParamElm, G4IonisParamMat, G4OpticalSurface: fixed problems reported by Coverity, mainly in assignment operator.
• Migrated code to new G4Exception scheme.

• Updated particle life time, mass and width according to PDG-2011.
• Added B-baryons, Bc mesons, Upsiron and Etac particles.
• Modified G4PrimaryParticle to avoid precision loss in conversion from momentum to energy.
• Fixed the scale of 'som0' in G4MuonRadiativeDecayChannelWithSpin and other internal quantities computations.
• Fixed bug in G4IsotopeMagneticMoment constructors.

• ASCII
  • Fixed case of dereferencing null pointer in G4tgbGeometryDumper DumpRotationMatrix() method.
  • Fixed bug on the usage of units. Addressing problem report #1240.
  • Migrated code to new G4Exception scheme.
• GDML
  • Extended GDML schema for materials to include ionisation potential (mean excitation energy) property to materials (GDML-3.0.1). Adapted reader and writer plugins accordingly.
  • Extended reader and writer to support new shape G4CutTubs.
  • Fixes in reader to properly consider units in import for G4GenericTrap, G4EllipticalTube, G4TessellatedSolid and G4Tet shapes.
  • Updated to comply with modification made in materials module.
• mctruth
  • Removed disabled obsolete HepMC code.

• New physics lists and builders:
  • Added OrderingParameterTable for G4PhysicsListHelper.
  • Added G4DNAEmPhysicsChemistry builder.
  • Added builders for anti-deuteron, anti-triton, anti-He and anti-Alpha.
  • Added builders for new LEND hadronic model. Enable to use LEND model and cross-section for low energy neutron in Shielding physics list (default is still neutron-HP) and added G4RadioactiveDecayPhysics. Updated G4PhysListFactory accordingly.
  • Added new builder G4IonFTFPBinaryCascadePhysics applicable for projectile ions up to 100 TeV and use it within QBBC; set transition energy to 2-4 GeV.
• Physics lists qualification changes (obsolete/partially disabled/experimental/supported):
  • Marked unsupported: QGSP_BERT_NOLEP, QGSP_BERT_TRV.
  • Marked supported: Shielding.
  • Marked experimental: CHIPS.
• EM physics builders and options:
  • Migrated EM standard, low energy and DNA builders to the new approach based on G4PhysicsListHelper: no more use of integer numbers to identify processes order; added de-excitation module to all EM builders.
  • Added G4BuilderType header with enumerator of builder types. Added type for all EM and Ion builders.
  • G4EmStandardPhysics_option3: experimental change of options: reduced low-energy limit to 10 eV; reduced StepFunction parameters per particle type; Activated Rayleigh Scattering. Use updated G4UrbanMscModel95 model for e+-.
  • G4EmLivermorePhysics, G4EmLivermorePolarizedPhysics, G4EmPenelopePhysics: use UrbanMscModel95 for e+-.
  • Modified G4EmExtraPhysics builder to use the new muon-nuclear model and process.
  • Set multiple-scattering polar angle limit from 0.2 to pi (reduction of number of steps for low-energy muons/hadrons) for G4EmStandardPhysics, G4EmStandardPhysics_option1, G4EmStandardPhysics_option2, G4EmStandardPhysics_option3, G4EmLivermorePhysics and G4EmPenelopePhysics.
  • G4EmStandardPhysics_option1,2: pi+-, kaon+-, proton, anti_proton use WentzelVI model of multiple-scattering.
  • Modified G4OpticalPhysics in builders; added new class G4OpticalProcessIndex. Addressing problem report #1216.
  • Modified G4OpticalPhysicsMessenger to allow G4State_PreInit for most commands.
  • G4EmStandardPhysics_option1: use UrbanMsc95. G4EmStandardPhysics_option2: same as Opt1 but with UrbanMsc93.
  • Added SetFiniteRiseTime() to G4OpticalPhysics and G4OpticalPhysicsMessenger.
  • Modified G4OpticalPhysicsMessenger so that commands setOpProcessVerbose and setTrackSecondariesFirst work in PreInit state.
  • Added process names in G4EmDNAPhysics. Changed elastic scattering model. Extended to the handling of HZE.
• Hadronic builders:
  • G4DecayPhysics, G4RadioactiveDecayPhysics: use G4PhysicsListHelper.
  • G4HadronElasticPhysics: added light anti-nuclei using LHEP model below 100 MeV and above using G4AntiNuclElastic model and G4ComponentAntiNuclNuclearXS.
  • G4HadronElasticPhysicsXS, G4HadronElasticPhysicsHP, G4HadronElasticPhysicsLEND: use G4HadronElasticPhysics for all particles and modified only neutron models and cross-sections.
  • Replaced G4UHadronElasticProcess by G4WHadronElasticProcess in G4HadronQElasticPhysics.
  • Removed obsolete G4HadronElasticPhysics93 configuration.
  • G4FTFPAntiBarionBuilder: added cross section to light anti-nuclei.
  • Fixed several small bugs in G4HyperonFTFPBuilder. Make use of new features of FTF in FTFP_BERT and in QGSP_FTFP_BERT: added builder G4HyperonFTFPBuilder for Hyperons using FTF and Bertini as models. Delete cross-section objects in G4FTFPAntiBarionBuilder.
  • HadronPhysicsFTFP_BERT: use G4HyperonFTFPBuilder, replacing G4MiscCHIPSBuilder; use G4FTFPAntiBarionBuilder for anti barions; FTFP_BERT and QGSP_FTFP_BERT uses new IonFTFPBinaryPhysics, replacing LHEPIonPhysics and IonBinaryPhysics, respectively.
  • G4HadronInelasticQBBC: use FTFP for all anti-baryons, set energy interval for FTFP above 2 GeV, below 6 GeV for Bertini.
  • Updated HadronPhysicsFTFP_BERT_TRV to use FTF for antibarions and hyperons, and to use CHIPS cross-sections for hyperons and kaons. This makes this configuration identical to HadronPhysicsFTFP_BERT, except for the transition energy. Changed overlap region between BERT/FTFP to be from 3 to 12 GeV, affecting FTFP_BERT_TRV physics-list.
  • Updated HadronPhysicsQGSP_FTFP_BERT to use FTFP for hyperons (G4HyperonFTFPBuilder).
  • Change in HadronPhysicsFTF_BIC to no longer use LEP for "misc" particles. Now using FTFP for antibarions (G4FTFPAntiBarionBuilder) and for hyperons (G4HyperonFTFPBuilder); also switch to CHIPS cross-section for kaons.
  • Modified G4HadronDElasticPhysics coherently with the changes made in G4DiffuseElastic.
  • Removed multiple implementations of FindInelasticProcess() method, now included in class PhysListUtil.
  • G4HadronElasticBuilder: use He3 x-section from the new AntiNuclei class to avoid crash in GHEISHA x-sections. G4HadronDElasticBuilder: x-sections from CHIPS, Diffuse model. G4HadronHElasticBuilder: alternative x-sections and models.
  • G4HadronElasticXS: use G4BGGHadronElasticXS proton cross-section.
  • G4HadronInelasticQBBC: use CHIPS x-sections for kaons; use Bertini up to 12 GeV, FTFP down to 3 GeV. Code cleanup in QBBC.
  • HadronPhysicsFTF_BIC: fixed bug for wrong deletion.
  • Removed He3 from elastic builders, because cross-section is still unavailable.
• Updates to lists:
  • Revised QGSP_FTFP_BERT to utilise the new FTF/P processes: changing them for anti-proton, anti-neutron; adding them also for the above list of light anti-ions (d,t,He3,alpha).
  • Added anti-triton to anti-barions. Use default energy limit for anti-barions, i.e. up o 100 TeV.
  • In QGSP_FTFP_BERT use G4IonBinaryCascadePhysics in place of G4IonPhysics.
  • QGSP_INCL_ABLA: enabled INCL/ABLA treatment for GenericIons.
  • Changed FTFP_BERT_TRV to use Fritiof for the nuclear capture of anti-protons at rest. Introduced new class G4QandFTFStoppingPhysics.
  • Added new physics list QGSP_INCLXX and assosiated builders. Uses INLC++ for protons, neutrons and pions in the energy range [0-3] GeV.
  • G4PhysListFactory: implemented selection of EM builder via physics-list name. EM variants (EMV, EMX, EMY, LIV, PEN) are now available only from physics-list factory, with the exception of FTFP_BERT_EMV and QGSP_BIC_EMY. The following now raise G4Exception explaining what is needed to do: FTFP_BERT_EMX, QGSP_BERT_EMX, QGSP_BERT_EMV.
  • FTF_BIC: removed use of LEP for Kaons; this required new builders G4KaonBuilder, G4FTFBinaryPionBuilder, G4FTFBinaryKaonBuilder, G4BinaryPionBuilder, G4BertiniKaonBuilder.
  • G4DataQuestionaire: updated for use more datasets. Updated lists to ask for check datasets needed: QGSP_BIC_HP, Shielding, QGSP_BERT_EMV, QGSP_INCL_ABLA.
• Fixed coverity issues in many files; most were uninitialised data members for processes/models.

• Introducing parallel layered mass geometry.
• Added new utility G4PhysicsListHelper for controlling process ordering in physics lists and related classes. If a process is added through the new method RegisterProcess(), the ordering parameters can be omitted. Default values for process ordering are applied. The user can choose to adopt his/her own parameter table by setting G4ORDPARAMTABLE.
• Default implementation of SetCuts() is provided. Now, it is no longer needed to implement a SetCuts() method (old scheme still supported). The obsolete method SetCutValueForOthers() has been removed. It is now allowed to set zero cut length in SetCuts commands.
• Added ReplacePhysics() in G4VModulerPhysicsList to replace physics constructor in a given physics list.
• Fixed bug in constructor of G4VUserPhysicsList reported by ATLAS for setting of DefaultCutValue flag.
• Changed G4VModularPhysicsList to issue messages always if physics is added and/or removed. Cleared compilation warning on gcc-4.6.
• Fixed potential case of deferencing null pointer in G4VModularPhysicsList::RemovePhysics().
• Removed unnecessary AtRest GPIL in G4ParallelWorldScoringProcess.
• Added banner for errors and warnings in G4ExceptionHandler. Redirect warnings to cout instead of cerr.

• Added GetSecondaryInCurrentStep() method in G4Step.
• Modified ParticleChange classes to properly handle local time change.
Addressing problem report #1224.
• Renamed GetVelocity() to CalculateVelocity() in G4Track; added accessor and modifier for velocity.
• Added G4VelocityTable CalculateVelocity() as singleton to avoid unnecessary creation of G4PhysicsVector objects in the event loop.
• Set velocity of track within UpdateStep() in G4ParticleChange.
• Modify G4Step::InitializeStep() to set correct velocity for optical photons.
• Modified G4ParticleChangeForTransport to fix a bug of wrong velocity in 'PostStepPoint' after Along/PostStepDoIt().
• Fix problems related to treatment of weight in G4ParticleChange. (addressing problem report #1243), to always set the parent track's weight to the value set by ProposeParentWeight() and ProposeWeight() methods for all particle change classes.
• Removed Conditionally-forced status in G4ForceCondition enum in G4SteppingManager, no longer applicable.
• G4RichTrajectoryPoint: added 'Pre/PostStepPointStatus' and 'Pre/PostStepPointWeight'. Fixed bug in copy constructor (volume touchable handle was not copied)

• Replaced G4cerr with G4cout and added verbosity control.
• Migrated to new G4Exception scheme.
• Fixed Coverity defects. Fixed warning on gcc-4.6.
• Management
  • Improvements to /vis/scene commands.
  • Added /vis/viewer/set/rotationStyle UI command.
  • Introduced RefreshIfRequired in viewer commands.
  • Added virtual function ResetView() in G4VViewer. Do not set auto-refresh for viewers that are not auto-refresh by default.
  • Added concept of default view parameters. These can be set in the visualization manager for future operations either by code or by command.
  • Added /vis/default/hiddenEdge and /vis/default/style.
  • Added printing of available colours on intialisation and in /vis/list.
  • Changed "Fatal" to "JustWarning" for non-existent model.
• Modeling
  • Relaxed some exceptions to warnings.
  • G4TrajectoryDrawerUtils: reworked TimesValidity.
  • G4PhysicalVolumeModel: added flag to G4PhysicalVolumeNodeID (true if drawn).
• Externals
  • Fixes to gl2ps.
• HepRep
  • Added /vis/heprep/scale and /vis/heprep/center commands.
• OpenGL
  • Fixed conflict with qgl.h and GL/gh.h in include files.
  • Added CMake support for Xm and Win32 components.
  • Fix in G4OpenGL header for proper inclusion of gl.h and glu.h for Qt build, when GL/X11 is not selected. Fixes compilation problem happening in either GNUMake and CMake builds, when selected Qt/GL driver on MacOS systems.
  • Fixed transparency problem.
  • Improved first pass rendering speed (reduced use of glFlush).
  • Fixed non-appearance of trajectories in Qt (added final glFlush).
  • Xm: fixed consistency and rest issue with control panels.
  • Stored mode: improved efficiency for handling transients (trajectories).
  • Fixed problem of missing trajectories drawing on Mac. Fixed slow drawing on Mac, by reducing the frequency of glFlush.
  • Fixed inconsistent parameters on Xm panels.
  • Use program loop to call display lists for "permanent objects" instead of using a "top display list".
  • Qt improvements. Bug fix for Qt < 4.4.
  • Fixed circles/squares problem.
  • Improvements in ogl/print and gl2ps.
• OpenInventor
  • Added support for CMake build of OpenInventor.
• RayTracer
  • Updated setup to use new GEANT4_USE_RAYTRACER_X11 variable.
• XXX
  • Implemented simple ad hoc tree and removed include/tree.

• Merged changes for migration to new G4ApplicationState.
• Fixes to port on OSX Lion and migrated to new G4Exception scheme.

• New data set G4NEUTRONXS-1.1 for evaluated neutron cross section data on natural composition of elements:
  • Evaluated cross section data derived from G4NDL-4.0.
• New data set G4PII-1.3 for shell ionisation cross-sections to be used in precise impact ionisation simulation.
  • Refined values of the theoretical cross-sections.
• New low-energy data set, G4EMLOW.6.23:
  • Added module penelope/bremsstrahlung for G4PenelopeBremsstrahlung v2008 model.
  • Updated FL2.dat and Paul data.
  • Added Bremsstrahlung data files into new sub-directory brem_SB. Data are published in S.M.Seltzer and M.J.Berger Atom. Data and Nucl. Data Tables 35 (1986) 345-418.
  • Added ecpssr Form Factor calculations from A. Taborda, M. Reis et al.
• New neutron data set with thermal cross sections, G4NDL.4.0:
  • Updated most data to ENDF/B-VII.
  • Added Zinc: Zn_nat.
  • Added Tantalum: Ta181:Ta-181 (neutron), Ta182:Ta-182 (neutron).
  • Replaced Lithium data: Li6, Li7.
  • Replaced Boron data: B10, B11.
  • All neutron data file below Elastic, Inelastic, Capture and Fission are replaced. New data comes from ENDF-VII.r0 library. New files are processed by NJOY-99.u364 with reconstruction tolerance of 0.001. Exceptions: data files for As75, Y90, Pa231, Pa233 and Th232 come from ENDF-VI.r8. No data files for Be7 and As74.
  • All excited isomer data are excluded, as currently not supported in Geant4.
  • All data file for isotopes heavier than Uranium are excluded.
  • Added simplified level scheme data of Np to Cf into inelastic/Gammas.
  • Total 389 isotopes including 3 natural abundance data.
• New photon-evaporation data set, G4PhotonEvaporation.2.2:
  • Updated z93.a237 element
• New radioactive decay data set, G4RadioactiveDecay.3.4:
  • New spontaneous fission capability and upgrade to forbidden beta decays.

• Migrated physics-list to use G4PhysListHelper.
• In DetectorConstruction in EM examples set I=78 eV (vs 75 eV)for Water. Fix in SteppingVerbose for proper treatment of 'OutOfWorld' cases.
• Archived advanced example 'Rich', no longer under maintenance.
• Various fixes, improvements, adoption of new features.
• Updated reference outputs.
• advanced
  • brachytherapy
    • Migrated to use new internal analysis module.
    • Deleted sensitive detector and ReadOutGeometry components.
  • ChargeExchangeMC
    • Implemented new framework for easy creation of various scene primitives.
    • Added new algorithms for energy deposit collection strategies.
    • Added new reconstruction parameter: expected momentum at the center of the target.
    • Optimized calculation of physical values. Code cleanup.
  • dnaphysics
    • New example exercising G4DNA processes.
  • eRosita
    • Imported old Livermore low-energy processes.
  • hadrontherapy
    • Added complete physics to the batch.mac file for batch runs.
    • Updates and corrections to macros and cleaned physics.
  • iort_therapy
    • New application specifically developed to address typical needs related to the Intra-Operative Radio-Therapy (IORT) tecnique.
  • human_phantom
    • Added macro batch.mac for running in batch mode.
  • microbeam
    • Removed obsolete hadron elastic builders. Removed usage of system commands.
  • microdosimetry
    • Switched to G4ionIonisation and 5 MeV proton.
    • Added new variables to histogram.
    • Added scoring of multiple-scattering.
    • Removed initStep test in SteppingAction.
    • Removed usage of system commands.
  • nanobeam
    • Switched to non-numbered physics list.
    • Removed usage of water material and cleanup.
    • Removed usage of system commands.
    • Removed one analysis option in HistoManager.
    • Updated README for use with full version of CLHEP.
  • xray_fluorescence
    • Migrated physicslist to standard EM+Fluo.
    • Added several classes (XrayFluoPhysListEmStandardFLUO, XrayFluoStepMax and XrayFluoStepMaxMessenger) and adapted code accordingly for use.
• basic
  • New set of basic examples B1-2-3-4 oriented to "novice" users and covering many basic general use-cases typical of an "application"-oriented kind of development.
• extended
  • analysis/A01
    • Corrected order of filling x and y for Cloud2D.
  • analysis/AnaEx02
    • New example showing the usage of histogram and ntuple manipulations using the Root analysis tool. All analysis manipulations (histo booking, filling, saving histos in a file, etc...) are located the class HistoManager.
  • common
    • First version of the placeholder of common classes to be adopted in the future extended examples.
  • electromagnetic/TestEm1,2,3,5
    • PhysicsList: set back default lowest production cut (990 eV).
    • Updated PhysListEmStandard for msc95 and de-excitation. Set lowest production cut to 100 eV in PhysicsList.
    • Updated SteppingVerbose.
    • Set msc95 as default in PhysListEmStandard.
  • electromagnetic/TestEm7
    • PhysicsList: set back default lowest production cut (990 eV).
    • Updated PhysListEmStandard for msc95 and de-excitation. Set lowest production cut to 100 eV in PhysicsList.
    • Updated SteppingVerbose.
    • Set msc95 as default in PhysListEmStandard.
    • Removed calls to obsolete methods and cleaned up options in PhysListEmStandardNR and PhysListEmStandardSS.
    • Added compiler depended selection of the math library to c2_function.
    • Allow tallies of different materials and sizes.
    • Added Water_1.05 material and test of variable density option in TestEm7.in macro and DetectorConstruction.
  • electromagnetic/TestEm9
    • Added extra run with 20 GeV mu- in TestEm9.in.
    • Set msc95 as default in PhysListEmStandard.
  • electromagnetic/TestEm11
    • PhysicsList: set back default lowest production cut (990 eV).
    • Updated PhysListEmStandard for msc95 and de-excitation. Set lowest production cut to 100 eV in PhysicsList.
    • Set msc95 as default in PhysListEmStandard.
    • Allow several absorbers to DetectorConstruction.
    • Modified RunAction, SteppingAction, SteppingVerbose, DetectorMessenger.
    • HistoManager and Messenger: removed 'csda' and 'stepMax' computation.
    • Updated all macros and Readme.
  • electromagnetic/TestEm12
    • PhysicsList: set back default lowest production cut (990 eV).
    • Updated PhysListEmStandard for msc95 and de-excitation. Set lowest production cut to 100 eV in PhysicsList.
    • Set msc95 as default in PhysListEmStandard.
    • Added G4EmDNAPhysics in PhysicsList and new macro dna.mac.
    • Reduced statistic in input macro.
    • Removed obsolete modifiers in PhysListEmStandard.
    • PhysListEmStandardGS, PhysListEmStandardWVI: low energy threshold for Coulomb scattering model reduced to 10 eV for e+- and to 100 eV for muons and hadrons.
  • electromagnetic/TestEm13
  • electromagnetic/TestEm14
    • Modified PhysListEmStandard, included AtomicDeexcitation, new G4PEEffectFluoModel and G4KleinNishinaModel.
  • electromagnetic/TestEm15
    • Updated PhysListEmStandard to multiple-scattering-95.
  • electromagnetic/TestEm18
    • Modified PhysListEmStandard, included AtomicDeexcitation, new G4PEEffectFluoModel and G4KleinNishinaModel.
    • Moved Fluo and Pixe tests to TestEm14.
  • field/field06
    • New example excercising gravity field.
  • geometry/g3tog4
    • Fixed vis.mac macros. Corrected README.
  • geometry/olap
    • Code cleanup.
  • geometry/transforms
    • New example demonstrating various ways of definition of 3D transformations for placing volumes.
  • hadronic/Hadr01
    • Added new builders for ions.
  • hadronic/Hadr02
    • New example providing simulation of ion beam interaction with different targets. Hadronic aspects of beam target interaction are demonstrating in the example including longitudinal profile of energy deposition, spectra of secondary particles, isotope production spectra.
  • medical/electronScattering
    • Updated PhysListEmStandard to multiple-scattering-95.
    • Added EGS_13MeV results.
    • Set msc95 as default in PhysListEmStandard.
    • RunAction: disabled storing of random seeds.
  • medical/electronScattering2
    • Corrected README comments on opt2 and opt3.
    • Updated PhysListEmStandard to match version in electronScattering.
    • Corrected physics-list setting in macros.
  • medical/fanoCavity[1,2]
    • Updated PhysListEmStandard_opt3 to multiple-scattering-95. Suppressed PhysListEmStandard_opt2, and use same configuration as for opt3 in PhysListEmStandard_GS, VWI, SS.
  • medical/GammaTherapy
    • Renamed step-limiter classes to avoid clash with class names in kernel.
  • optical/LXe
    • Exercise the optics_engine in input macro.
  • parameterisations/gflash
    • Fixed bug in ExGflashSensitiveDetector::Initialize() for creation of new hits collection.
  • persistency/gdml
    • G02: Updated test.gdml to add example of ionisation potential to materials. Requires upgrade of the schema to GDML-3.0.1.
    • G03: Changed from no longer supported physics-list QGSP-BERT-EMV to QGSP-BERT.
  • radioactivedecay/rdecay01
    • Fixed bug in time counting in RunAction.
    • Compute activity from primary ion in RunAction.
    • Modified setting for total kinetic energy in histo 6.
    • Suppressed histo #9.
  • radioactivedecay/rdecay02
    • Migrated physics list to remove use of obsolete low energy EM processes.
    • Corrected pre-processor lines in exrdmHisto.
    • Added macro files and user data files to illustrate the possibility for a user to define its own radioactive decay data and evaporation data. An example No252.g4mac shows also how to treat nuclei with Z>100.
    • Added commands '/process/em/fluo true' and '/process/em/auger true' in macro files to have the atomic relaxation taking place, with the new EM interface for atomic relaxation. Updated README.
  • runAndEvent/RE04
    • New example demonstrating how to define a layered mass geometry in a parallel world.
• novice
  • N03
    • Get pointers to UserAction classes via G4RunManager.
  • N04
    • Optimised initialisation of Phi angles in parameterisations to avoid recomputation of trigonometry.
  • N06
    • Use Spline interpolation for some property-vectors.